IMO, the "best" structure depends on your use case. If you’re feeding data to big platforms (Google, etc.), go compacted with schema.org. If it’s internal systems, maybe custom vocabs are fine.
Tools-wise, jsonld.js (https://github.com/digitalbazaar/jsonld.js) is solid for processing. And yeah, avoid the YOLO approach—your future self (or the next dev) will hate you.
Got burned once with over-nested data. Never again.
Tools-wise, jsonld.js (https://github.com/digitalbazaar/jsonld.js) is solid for processing. And yeah, avoid the YOLO approach—your future self (or the next dev) will hate you.
Got burned once with over-nested data. Never again.